Discovering insights from scientific literature using NLP

Introduction

• In biomedical sciences, some of the most valuable data is stored in unstructured text. Even with search engines, humans still need to read through entire articles to find the information they need.
• This is where Natural language processing (NLP) can help with a wide variety of techniques used to process data generated from human language.
• But NLP alone is not enough, by combining structured and unstructured data and organizing the insights into Knowledge graphs allow us to derive meaning and relationships.
• This allows scientists to navigate the relationships between compounds, protein, symptoms, and other biomedical concepts without needing to read and re-read countless articles. This also means, that as these scientists work on the data, finding new connections, we will be able to better infer new relationships.

The Webinar will show how you can use NLP to build a biomedical knowledge graph and how this will be helpful in your research.

What will you learn?

• How natural language processing (NLP) and/or text mining can extract - data from scientific papers
• How to build a scalable pipeline for processing a large corpus of data
• How to combine entities extracted from text with structured data to create a knowledge graph
• How to use a knowledge graph to help researchers
• How to bootstrap a NER model
• How to use topic modeling to help users browse a corpus

Presenters:

Alex Thomas
https://www.linkedin.com/in/alnith/
Alex Thomas is a principal data scientist at Wisecube. He is the author of the newly published O’Reilly book ‘Natural Language Processing with Spark NLP’ and has used natural language processing and machine learning with clinical data, identity data, employer and job seeker data, and now biochemical data. Alex has used Apache Spark since version 0.9 and has worked with NLP libraries and frameworks including UIMA and OpenNLP.